Multilayer Perceptron Neural Network Approach for Voltage Sags Detection and Classification

This paper represents the analysis of Voltage Sags detection and classification based on S- Transform by using half-cycle windowing technique (HCWT) with Multilayer Perceptron Neural Network (MLPNN) that had been conducted. The main objectives are to analyze the detection and classification of volta...

Full description

Saved in:
Bibliographic Details
Published in2021 6th IEEE International Conference on Recent Advances and Innovations in Engineering (ICRAIE) Vol. 6; pp. 1 - 5
Main Authors Anak Tarip, Anthony Roston, Daud, Kamarulazhar, Ismail, Ahmad Puad, Tajudin, Aimi Idzwan, Omar, Saodah, Mohd Hussain, Mohd Najib
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.12.2021
Subjects
MATLAB
multilayer perceptron neural network
Multilayer perceptrons
Neural networks
Power quality
Programming
s-transform
Technological innovation
Transforms
Voltage fluctuations
voltage sags
Online AccessGet full text

Cover

More Information
Summary:This paper represents the analysis of Voltage Sags detection and classification based on S- Transform by using half-cycle windowing technique (HCWT) with Multilayer Perceptron Neural Network (MLPNN) that had been conducted. The main objectives are to analyze the detection and classification of voltage sag i.e. pure sag and sag with noise based on S-Transform (ST) by using HCWT with Neural Network (NN). Besides, comparison between half-cycle and one-cycle windowing technique (OCWT) also made for analysis. Both types of sag signals were generated by using equations with programming in MATLAB. The data produced were then extracted using HCWT based on ST to be input of Neural Network for the classification process. The accuracy of classifications was displayed in percentage. It was verified that the OCWT is much better than HCWT since it has stable and smooth detection line. Besides, classification results for OCWT is much better than HCWT. It is also concluded that ST is able to detect and classify sag signal even in noisy condition.
DOI:10.1109/ICRAIE52900.2021.9703832